UNISEP: A Unified Sensor Placement Framework for Human Motion Capture and Wearables

The proliferation of wearable sensors and monitoring technologies has created a need for standardized sensor placement protocols. While existing standards like the Surface Electromyography for Non-Invasive Assessment of Muscles (SENIAM) recommendations for electromyography (EMG) and the 10-20 system for electroencephalography (EEG) address modality-specific applications, no comprehensive framework spans different sensing modalities and applications. We present the Unified Sensor Placement (UNISEP) framework to facilitate reproducible handling of human movement and physiological data across various systems and research domains. The framework provides a method to describe coordinate systems and placement protocols based on anatomical landmarks, and is designed to complement existing data-sharing standards such as the Brain Imaging Data Structure (BIDS) and Hierarchical Event Descriptors (HED). Even during its proposal stage, the UNISEP approach has been adopted by the EMG-BIDS extension (BIDS version 1.11.0), confirming the community need for a unified, machine-readable sensor placement framework. The UNISEP framework facilitates consistency, reproducibility, and interoperability in applications ranging from lab-based clinical biomechanics to continuous health monitoring in everyday life.

💡 Research Summary

The paper addresses a critical gap in human movement and physiological monitoring: while modality‑specific placement standards such as SENIAM (EMG), Mason‑Likar (ECG) and the 10‑20 system (EEG) exist, there is no unified framework that can describe sensor locations across different technologies in a machine‑readable way. To fill this void, the authors propose UNISEP (Unified Sensor Placement), a technology‑agnostic schema that defines anatomical coordinate systems for each body segment using reliable, palpable landmarks. Each coordinate system is specified by an origin and three orthogonal axes aligned with functional anatomical directions, following the International Society of Biomechanics (ISB) conventions.

Sensor placement is expressed in three steps: (1) identify the relevant body segment and its anatomical coordinate system; (2) locate the sensor within that system using normalized percentages (0–100 %) along each axis, which automatically scales to individual body proportions; and (3) record the measurement method (visual inspection, tape, 3D scan, etc.) and any reference standard (e.g., SENIAM, 10‑20). This normalized representation enables the same description to be applied to subjects of different sizes and to multiple sensor modalities (EMG electrodes, inertial measurement units, motion‑capture markers) placed on the same segment.

A hierarchical coordinate‑system concept further allows complex sensor arrays—such as high‑density sEMG grids—to be described with a parent anatomical system (e.g., thigh) and a child local system (grid coordinates in millimetres). The relationship is captured through an “AnchorElectrode” field that links the child system to a specific electrode’s normalized position. The authors demonstrate this approach in the EMG‑BIDS extension (part of BIDS version 1.11.0), showing how multi‑device recordings can be annotated with JSON side‑car files that contain both normalized body‑relative positions and precise intra‑grid layouts.

UNISEP is deliberately designed to complement, not replace, existing standards. It provides a common spatial language that can encode the recommendations of SENIAM, Mason‑Likar, 10‑20, etc., within a single, interoperable metadata structure. By integrating with BIDS and HED, UNISEP supplies the machine‑readable metadata required for automated pipelines, cross‑study comparisons, and FAIR data principles (Findability, Accessibility, Interoperability, Reusability).

The paper acknowledges a current limitation: sensor orientation is not yet formalized, which is essential for devices like IMUs where axis alignment matters. The authors suggest future extensions to incorporate orientation descriptors and dynamic landmark displacement corrections.

In summary, UNISEP offers a comprehensive, anatomy‑based coordinate framework that standardizes sensor placement across EMG, ECG, EEG, IMU, optical motion capture, and emerging wearables. Its adoption by EMG‑BIDS illustrates immediate community relevance, and its compatibility with major data‑sharing standards positions it as a foundational tool for reproducible, interoperable human movement research and continuous health monitoring.